flyte
Add your description here
Decision gist · record as of 2026-08-14
Yes, with conditions. Flyte 2 is actively maintained (released 2 days ago) and offers a cohesive Python-first approach to ML orchestration with low install friction. However, the license treatment is unclear in the metadata despite Apache 2.0 being mentioned in the description—verify the actual license before production use. The Rust controller is marked experimental and under rapid development with known gaps. Best suited for teams building ML pipelines who can tolerate a large dependency footprint and want async-native workflows.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or later
- Low friction installation as a pure Python wheel.
- Active maintenance with a release 2 days ago.
License · maintenance · safety
(unclear)
last release 2026-08-12 (2 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 173,423 downloads/mo, #10,309 on PyPI
Alternatives
Verify before relying
pip install flyte
import flyte
env = flyte.TaskEnvironment(
name="hello_world",
image=flyte.Image.from_debian_base(python_version=(3, 12)),
)
@env.task(retries=3, cache="auto")
async def predict(x: int) -> int:
return 2 * x + 5- License classification is marked unclear in the fact sheet; Apache 2.0 is mentioned in the description excerpt but not confirmed in metadata
- Whether the Rust controller (flyte_controller_base) is required for typical use or truly optional as described
- Production readiness status beyond 'generally available' announcement
What it is and what it does
Flyte is a Python-native orchestration framework for building and running ML pipelines, models, and agents. It lets you define tasks as decorated async or synchronous Python functions, compose them into workflows, and execute them locally during development or remotely on Kubernetes at scale. The SDK handles task parallelism, retries, caching, and resource management automatically.
The framework supports multiple execution modes: local development with a CLI and optional TUI, a Devbox environment for testing against a local Kubernetes cluster, and remote execution on production infrastructure. It also includes app serving capabilities for deploying models via FastAPI. With 23 runtime dependencies including async-lru, pydantic, protobuf, and cloud storage support, it integrates deeply with modern Python async patterns and cloud ecosystems.
Use it for
- Define and run multi-step ML training pipelines with automatic task parallelism and fault tolerance.
- Serve trained models as HTTP endpoints using FastAPI integration within the same framework.
- Develop workflows locally with the CLI and TUI, then deploy to Kubernetes without code changes.
- Build data processing workflows that combine async task execution with built-in caching and retries.
- Orchestrate agent workflows that require fine-grained control over concurrency and resource allocation.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, with conditions.
Flyte 2 is actively maintained (released 2 days ago) and offers a cohesive Python-first approach to ML orchestration with low install friction. However, the license treatment is unclear in the metadata despite Apache 2.0 being mentioned in the description—verify the actual license before production use. The Rust controller is marked experimental and under rapid development with known gaps. Best suited for teams building ML pipelines who can tolerate a large dependency footprint and want async-native workflows.
Install
flyte on PyPI
Before you install
Low friction installation as a pure Python wheel. Active maintenance with a release 2 days ago. Requires Python 3.10 or later and brings in 23 runtime dependencies including async utilities, protobuf, pydantic, and cloud storage support.
Requires Python 3.10 or later
Quickstart
pip install flyte
import flyte
env = flyte.TaskEnvironment(
name="hello_world",
image=flyte.Image.from_debian_base(python_version=(3, 12)),
)
@env.task(retries=3, cache="auto")
async def predict(x: int) -> int:
return 2 * x + 5Verify before relying
- License classification is marked unclear in the fact sheet; Apache 2.0 is mentioned in the description excerpt but not confirmed in metadata
- Whether the Rust controller (flyte_controller_base) is required for typical use or truly optional as described
- Production readiness status beyond 'generally available' announcement
Package facts
| License | Not declared unclear |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 23 packagesaiofilesclickcloudpickledocstring_parserfsspecconnectrpcobstoreprotobufpydanticpyyamlrich-clickhttpxkeyringmsgpacktomlasync-lrumashumaroaiolimiterflyteidl2packagingsentry-sdkpyOpenSSLasyncssh |
| Maintenance | Actively maintained 2 days since the last release |
| First released | |
| Downloads | 173,423 / month, #10,309 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: flyte-2.6.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “async python workflows”
- flyteFlyte is a Python SDK for defining, orchestrating, and executing ML…
- llama-index-workflowsAn async-first, event-driven framework for orchestrating multi-step…
- temporalioTemporal Python SDK provides a framework for building durable,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Distributed Computing packages
gRPC Python is an HTTP/2-based RPC framework that enables you to define and call remote procedures across network boundaries using protocol buffers for serialization.
Install it if you need RPC communication in a distributed system or are integrating with existing gRPC services.
execnet lets you spawn and communicate with Python interpreters across local processes, remote hosts, and different platforms, using a simple API for task distribution and inter-process messaging.
However, the aging maintenance status (275 days since last release) means you should verify it meets your concurrency and performance needs before committing to a…
Cloudpickle extends Python's standard pickle module to serialize lambda functions, interactively-defined functions and classes, and other constructs that the default pickle cannot handle, making it suitable for cluster computing and remote code execution.
Install it if you need to serialize lambda functions, interactively-defined code, or non-standard Python constructs for cluster computing or distributed execution.
Provides a unified, open()-compatible Python API for streaming large files from remote storage (S3, GCS, Azure, HDFS, SFTP, HTTP) and local filesystems, with transparent compression support.
Install it if you work with large files on cloud storage or remote systems and want to avoid writing boilerplate around multiple SDKs.
Portalocker provides cross-platform file locking with support for exclusive and shared locks, plus Redis-based distributed locks and process-aware PID file locking.
Install it if you need file or process coordination; the optional extras (pywin32, redis) are only required for specific lock types.
Ray is a distributed computing framework that scales Python applications from a single machine to multi-node clusters, providing abstractions for parallel tasks, stateful actors, and shared objects.
See also flytekit · flyteidl2 · flyteidl · metaflow · kfp-pipeline-spec · skypilot-nightly · ob-metaflow · kfp-server-api · zenml · azureml-pipeline-core